Data entry ops ·

uProc saves time and engineering resources by using n8n to scrape banking data from a multi-page website

The problem

Collecting banking reference data (Swift codes) from a multi-page website was challenging because the data was scattered across sources in different formats and sometimes outdated. The prior Python/Scrapy approach required repetitive manual coding work — selecting HTML tags, formatting, and processing — making it time-consuming.

First attempt

Python scripts using Scrapy were technically adequate but required extensive repetitive manual coding — selecting tags, formatting, and processing data — making the approach impractical to maintain.

Workflow diagram · grounded in source
1
Cache directory initialization
trigger
“the Execute Command node (to automatically create a local cache directory before starting the web-scraping process and avoid scraping the same pages)”
2
HTTP page fetch
integration
“the HTTP Request node (to access data from the https://www.theswiftcodes.com website)”
3
HTML content extraction
integration
“the HTML Extract node (to extract the desired content from the website based on their HTML tags)”
4
Deduplication routing
routing
“the IF node (to filter information based on conditional logic, for example, checking whether a Swift code already exists in the database)”
5
Store data in MongoDB
output
“stores the collected information in MongoDB”
Reported outcome

Miquel replaced the Python scripts with a 22-node low-code n8n workflow that scrapes all country pages on theswiftcodes.com and stores the data in MongoDB, saving time and engineering resources by automating away repetitive coding.

Reported metrics
Time and engineering resourcessave precious time and resources
Reported stack
n8nMongoDBuProcScrapy
Source
https://n8n.io/case-studies/uproc/
Read source ↗

Frequently asked questions

What did this team achieve with this AI workflow?

Miquel replaced the Python scripts with a 22-node low-code n8n workflow that scrapes all country pages on theswiftcodes.com and stores the data in MongoDB, saving time and engineering resources by automating away repe…

What tools did this team use?

n8n, MongoDB, uProc, Scrapy.

What results were reported?

Time and engineering resources: save precious time and resources (source-reported, not independently verified).

What failed first in this deployment?

Python scripts using Scrapy were technically adequate but required extensive repetitive manual coding — selecting tags, formatting, and processing data — making the approach impractical to maintain.

How is this data entry ops AI workflow structured?

Cache directory initialization → HTTP page fetch → HTML content extraction → Deduplication routing → Store data in MongoDB.

Related data entry ops cases